Baseball Postseason Predictions Many baseball fans have a love affair with two things: their favorite team and statistics. Bruce Bukiet, an associate professor of mathematical sciences, shares his predictions and mathematical models for this year's Major League Baseball playoff standings.

#### Baseball Postseason Predictions

• Download
• `<iframe src="https://www.npr.org/player/embed/232159376/232159367" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">`
• Transcript
Baseball Postseason Predictions

# Baseball Postseason Predictions

#### Baseball Postseason Predictions

• Download
• `<iframe src="https://www.npr.org/player/embed/232159376/232159367" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">`
• Transcript

Many baseball fans have a love affair with two things: their favorite team and statistics. Bruce Bukiet, an associate professor of mathematical sciences, shares his predictions and mathematical models for this year's Major League Baseball playoff standings.

(SOUNDBITE OF BASEBALL BEING HIT)

(SOUNDBITE OF CHEERS)

JOHN DANKOSKY, HOST:

Ah, yes. Baseball's final four is finally set. In the American League, the Boston Red Sox will take on the Detroit Tigers in the ALCS. In the National League, the L.A. Dodgers will face the St. Louis Cardinals, with the winners of those series meeting in the World Series less than two weeks away.

Now, baseball has always been a numbers game but the RBIs and ERAs of my youth have been taken over by FIP and WAR, that's W-A-R, that make up a new kind of advanced baseball analysis called sabermetrics. Now for the past 13 years my next guest hasn't been using sabermetrics, but what he says is just simple math, to predict what teams will make it to the top. Bruce Bukiet is here to tell us about his predictions and math models.

He's associate professor of Mathematical Science at the New Jersey Institute for Technology. Welcome, Bruce.

BRUCE BUKIET: Thank you for having me.

DANKOSKY: So how did you get into baseball predictions?

BUKIET: Well, I was reading some articles one day in the library and I saw one on using math to compute tennis and to understand tennis. And I loved when I could connect math with relevant issues in daily life. And I thought, I'm a baseball fan. I would love to see what I could do to understand baseball better. And so I started working on that.

DANKOSKY: So when you started working on it what statistics were you putting in your model? What were you looking at?

BUKIET: Right. So at the beginning I started with just a very simple kind of brute force approach. In those days, in the late '80s, it was easy to get data on how many walks, singles, doubles, triples, homers, and outs each player got. And so I just started modeling that. If the first player gets one of those six things and then if the next player gets it, you know, one of those six, what would happen?

And to go from there and realizing that the order matters in baseball. If you get a single, then a homerun, that's two runs. Whereas, if you get a homerun and then a single, that's only one run. So since I had to deal with all of the possible orders that things could happen and there are around 40 players that bat in a game for a single lineup, it quickly led me to realize as my computer slowed down, that it would take many millions of years to understand even one batting order.

DANKOSKY: I'm John Dankosky and this is SCIENCE FRIDAY from NPR. And we're talking about baseball and math on the program. Now, look, what you're talking about does seem fairly simple. There's all this advanced baseball statistical analysis out there - sabermetrics, as I said before. They're trying to look at models that they say is more predictive of future behavior. Do you go in for any of this stuff? Do you look at any of these sabermetrics?

BUKIET: Well, I think that there's certainly something there. I am not sophisticated enough, nor am I a baseball expert, so I used test data for all the individual players and try to go from there and assume that players will perform similarly to the way that they've performed over the last several years. At some point maybe I'll be able to add some of that analysis into my modeling.

DANKOSKY: So how good have your predictions been?

BUKIET: Well, this year was a particularly good year. I was able to see how I did against so-called experts comparing it with people from Yahoo! Sports, ESPN, Sports Illustrated. And I was able to pick a number of the teams that made the playoffs, about seven out of the 10. So that doesn't sound so great but out of the 64 other prognosticators, only seven of them got seven right. So I was up there with them.

And then in terms of the number of games, my model enables us to pick how many games - to see how many games each team should win and what order the teams should come in in, and most people don't put that data out or their predictions out, but of about the half dozen there I came in, I think, first in almost all the categories that - it was just a fortuitous season.

But the model does seem to have a good result. So I've been doing that for a number of years.

DANKOSKY: But of course there's always some outliers. You didn't pick my Pittsburgh Pirates to make the playoffs.

BUKIET: That right.

DANKOSKY: In part, because they've not done well over the course of the last couple of years, so your model says they probably weren't going to do well again this year.

BUKIET: Right. And so those individual players, most of the players I guess were the same at the beginning of the year - and there's a lot of things that you would think in baseball how can you do this at the beginning of the year? You don't even know who's going to play.

My favorite team is the Mets and I think at the end of the season they maybe had two or three of the starters they had on opening day starting for them. So it's amazing how many things kind of even out on many teams. The substitute players probably around as good as the player that's playing on a daily basis or the team will go out and get someone with similar ability.

But there's always a few who mesh up this year. The Pittsburgh Pirates, as you said, they were the ones that I was up by 21 games. That was the worst. But on the other hand, there were four that came out exactly right, which also is really high. And 18 out of 30 teams were within five games or less.

DANKOSKY: See, now we've got these final four teams here. First of all, did you pick these four teams to be in the place where they are right now?

BUKIET: Well, I had three of them. I did not have the Boston Red Sox.

DANKOSKY: OK. You didn't have the Red Sox. Are you able to tell us now who's going to win the World Series? Do you know?

BUKIET: I can't tell you the World Series but I can tell you for the league championship series what the odds are. We have that the Red Sox should be favored to win against the Tigers at a 57 percent chance of winning. But the Dodgers have a much better 68 percent chance of overcoming the Cardinals. And so what you have is the most likely outcome, 39 percent, is of a Boston-Los Angeles World Series.

DANKOSKY: So you've just made some people very, very happy in Boston and Los Angeles. Well, good luck with your predictions and good luck to my Pirates next year. Maybe you will get it right this time, Bruce.

(LAUGHTER)

BUKIET: Maybe. Thank you so much for having me.

DANKOSKY: And good luck to your Mets as well. Bruce Bukiet is associate professor of Mathematical Science as the New Jersey Institute for Technology.

Copyright © 2013 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.